Spatial and Temporal Distribution of Root Water Uptake 8 of an Almond Tree under Microsprinkler Irrigation 7 2 - 7 6 Kouman S. Koumanov1, Jan W. Hopmans2, and Larry W. Schwankl2 2 p . pm 1 Fruit Growing Institute, Plovdiv, Bulgaria e 4, k.co7-3 2 Dept of LAWR, University of California, Davis, U.S.A. un2 si0 . 24, Isringerl-005-0 ATinh btehs estp rraaotociatt l z aonnde toefm ap soirxa-ly peaartt eorlnd mofi croroo-ts pwraintekrl eurp itrarkigea itned p aalrmtiaolnlyd wtreeett.e Td hseo iwl awtears bsatuladniecde lp1 os7 of about one quarter of the root zone’s wetted soil volume (2.0 x 2.0 x 0.9 m) was Vw.2 0 determined by catch canes, neutron probe and tensiometer measurements. Twenty-five 6,w0 0ws neutron probe access tubes with catch canes were distributed in a square grid of 50 cm ay 20le at 1007/ rsepgaucilnarg . pEaitgtehrtn pabiertsw oefe nt enthsieo macecteersss wtuebree si.n sNtaeluletdro na t pdreopbteh s roefa d8in2g.5s aantd 1957 .5c mc md einp tha Mb. 0 increments and tensiometer readings were taken at time intervals of 4 to 24 hours. The rate a e, aile/1 of soil water depletion was calculated and used to estimate the spatial and temporal cvl nac distributions of root water uptake. Soil water dynamics was studied in two stages: 1) ies rti during a week of conventional irrigation management with three irrigation events; and 2) cia n Son m/ during a period of 16 days without irrigation, after the monitored soil volume was ioatico thoroughly moistened so that soil water was easily available everywhere, initially. The atcr. zones of maximum root water uptake were the same for both stages in periods of high glie ibg local rates of water application. Hence, the almond tree appeared capable to redirect its run n Irl ppri root activity towards regions of the most favorable water regime with minimum soil water ias stress. After water applications, root water uptake occurred initially near the tree trunk and n. d k egin then progressed towards the root system periphery, thereby changing locations of hii isor//l maximum root water uptake and shifting to root zone regions with minimum soil water ble p: stress. uht PTht Key words: root water uptake, microsprinkling, water balance, fruit trees. Introduction Proper water and soil management in orchards is essential for both sustainable agriculture and integrated food production (Cross and Dickler, 1994). As a result, there is increasing interest in microirrigation systems that can efficiently supply plants with both water and fertilizers. Farmers can use microirrigation to meet plant water requirements, providing an optimal soil water regime in the root zone (Bucks and Davis, 1986), thereby optimizing soil water availability during various crop growth stages using water stress to control growth and fruiting (Behboudian et al., 1998; Chalmers et al., 1984; Goldhamer, 1996; Lampinen et al., 1995; Proebsting et al., 1989). In addition, with proper irrigation management, one can regulate water and solute flows in the root zone to prevent both leaching of agrochemicals to ground waters and high salt concentrations in the soil, including by fertigation. However, in contrast to an intensive technical improvement of microirrigation systems during the last decades, methods for irrigation scheduling have not been essentially changed since the appearance of microirrigation in the sixties of the past century. 1 Irrigation scheduling is often based on the water balance method. However, when using localized irrigation, it is difficult to evaluate several components of the water balance as complications arise from preferential root development in the wetted soil volume (Sakovich and Post, 1986; Bielorai, 1985) and from non-uniform water uptake within root zone (Kramer and Boyer, 1995). In fact, most root water uptake from localized systems may be isolated to the irrigated regions of the root system, even when water is temporarily available throughout the entire root zone after precipitation (Kramer, 1950). Therefore, efficient water management in micro-irrigated rooting systems depends on knowledge of the spatial and temporal distribution of root water uptake, as well as on the ability to predict variations of soil water status in the root zone after irrigation. Whereas there are some investigations on the root morphology of trees, including the spatial distribution of roots under localized water application (Dasberg et al., 1985; Kjelgren et al., 1985; Meyer and Peck, 1985; Roth and Gardner, 1985; Sakovich and Post, 1986; Clausnitzer and Hopmans, 1994), information to date on the spatial and temporal distribution of root water and nutrient uptake is limited, especially for partial-wetted soils (Clothier, 1989; Kramer and Boyer, 1995). Root water uptake models that can describe spatial and temporal patterns were developed by Coelho and Or (1996) and Vrugt et al (2001). To provide experimental data for the development, verification and calibration of root water uptake models, thereby improving microirrigation scheduling and management, a field study was initiated in a micro-sprinkler irrigated almond orchard. This study was part of a larger project for evaluation of the physical performance of various microirrigation systems in orchards (Schwankl et al., 1996). The objective of the present study is to analyze the spatial and temporal root water uptake patterns of a micro-sprinkler irrigated almond tree. The daily water balance, water application uniformity and application efficiency of this micro-sprinkler system were discussed in a previous paper by Koumanov et al. (1997). Material and methods The investigation was conducted during the summer of 1995 in a six-year old almond orchard (Nickels Ranch) located 90 km north of Davis, California in the Sacramento Valley. The almond trees (Prunus amigdalus) in the experimental plot were of the “Butte” variety, grafted on “Lovel Peach” rootstock. Tree spacing was 4.8 x 6.6 m and the canopy coverage of the soil surface was 60%, determined by the area of tree-crown vertical projection and averaged for 30 trees. The soil surface was maintained free of weeds by periodic herbicide treatment. The soil had not been tilled since tree planting in 1990. The surface terrain is irregular and slightly undulated. The majority of soils in the orchard were classified as the Arbuckle series (Harradine, 1948) with young alluvial deposits of at least 0.60 m overlaying an old clay deposit within 1.50 m from the soil surface. The A-horizon consists of a 0.30-0.4 m thick gravely sandy loam surface layer. The B-horizon is a gravely loam to a depth of 0.90-1.50 m. Soil porosity decreases with depth and internal drainage is restricted by a clayey layer below the B-horizon. Table 1 lists particle size distribution, bulk density, field capacity (FC), and volumetric gravel content of the various soil layers. Additional information on the soil data is presented in Koumanov et al. (1997). Field capacity was determined independently from daily neutron probe measurements after a ponding infiltration test was done between the tree rows near the experimental plot. 2 Micro-sprinkler irrigation was applied on about one third of an 8.8 ha almond orchard. The micro-sprinklers (Bowsmith Fan-jet1) were of the fixed head type), producing 22 single streams of water in a full circle wetting pattern. At an operating pressure of 0.15 MPa, the micro-sprinkler average discharge was about 41.7 L h-1 over an effective wetting radius of approximately 2.0 m. During the first year after tree establishment, the micro-sprinklers were placed near the trees (one per tree) and water application was limited to a wetting circle of 0.50 m diameter by mounting throw limiters (Top HatTM) on the microsprinkler heads. After the first year, the micro-sprinklers were placed midway between trees in the tree row. According to Willoughby and Cockroft (1974), few months is enough for fruit trees to redirect their root activity in response to any change in the soil-wetting pattern. Hence, the root systems of the almond trees probably were able to readjust to the changing water application pattern. The study objectives were addressed using detailed soil water monitoring in the root zone of a single representative almond tree. The experimental plot covered about one quarter of the wetted area of one micro-sprinkler (Fig. 1). The microsprinkler was located at the origin of the coordinate system, with the stem of the almond tree 2.40 m away from the micro-sprinkler. In the 2.0 x 2.0 m monitored area, 25 PVC neutron probe access tubes (diameter 50 mm) were installed in a square grid using a separation distance of 50 cm, to a depth of 1.2 m. Despite that many other soil moisture devices are readily available, such as TDR and soil gypsum blocks, these are mostly considered to be point measurements. The neutron probe is among the most accurate techniques to determine representative soil water content values for plot-size soil domains that are as large as our experimental study area. In addition, eight pairs of tensiometers were installed in a regular pattern between the access tubes at depths of 82.5 cm and 97.5 cm, respectively. Each neutron probe access tube was assumed to represent a 0.50 x 0.50 m (0.25 m2) area. In the water balance calculations, however, the access tubes along the western and the northern boundaries of the experimental plot were assumed to represent only half of the experimental area, i.e. 0.50 x 0.25 m. Thus, the experimental plot approximated one quarter of the wetted area of a single almond tree, i.e. 2.25 x 2.25 m (5.0625 m2). Since the water application rates were low and no water ponding occurred, there was no evidence of the instruments affecting the soil water regime, despite that a large number of neutron probe access tubes and tensiometers were concentrated in a relatively small area. The neutron probe (503 Hydroprobe2) was calibrated from gravimetric measurements using undisturbed soil samples collected at depths of 15, 30, 45, 60, 75, and 90 cm during the access tube installation, with additional soil cores collected near an access tube from a nearby experimental plot. Separate calibration curves, representing soil water content versus count ratio, were used for the 15 cm surface soil, the 30-60 cm soil depth interval and the 75-90 cm soil depth (Koumanov et al. 1997). The count ratio is defined as the ratio of actual neutron count and standard count, thereby allowing for correction of systematic changes in neutron production by the radioactive source of the neutron probe. Standard counts were obtained as averages of series of 30 successive counts taken before each measurement in the vicinity of the experimental plot with the neutron probe placed on the top of its case. Standard errors of volumetric water content using these three calibration curves were approximately 0.01 cm3 cm-3 for the 0-15 cm soil depth interval and 0.02 cm3 cm-3 for all other measurement depths. Soil water matric head measurements were taken using a pressure transducer (Tensimeter3). After linear interpolation of the tensiometer 1 Bowsmith, P.O.Box 428, Exeter, CA 93221, USA. 2 Campbell Pacific Nuclear Corporation, Martinez, CA, USA. 3 Soil Moisture Systems, Las Cruces, NM, USA. 3 data, vertical water fluxes across the lower boundary (90 cm depth) were evaluated for all neutron probe access tube locations using Darcy’s equation: ( )dH ( ) q = - K q 1 dz where q denotes the water flux density (cm h-1), dH/dz is the total head gradient (cm cm-1), and K(q) denotes the unsaturated hydraulic conductivity (cm h-1). Unsaturated hydraulic conductivity data were computed from the expression (Mualem, 1976; van Genuchten, 1980): ( ) ( ) K(q )= K Sl 1- 1- Sl/m m 2 ( ) s e e 2 where S is defined by the van Genuchten soil water retention expression e q - q ( ) l ( ) r =S = 1+(- a .h)n - m m=1- 3 q - q e n s r and q is the saturated water content (cm3 cm-3), q is the residual water content (cm3 cm-3), s r K is the saturated hydraulic conductivity (cm h-1), h is the soil matric potential head (cm), s and a (cm-1), n, l are empirical fitting parameters. Soil hydraulic parameters were taken from Andreu et al. (1997) and Vrugt (2001), with values of q = 0.281, q =0.108, a = s r 0.153, n = 1.612, l = 0.5, and K = 0.02 cm h-1. s For water flux estimations from the bottom of the root zone, we used the unsaturated hydraulic conductivity function for the 90-cm soil depth, using parameter values reported by Andreu et al. (1997). The h-values in Equation (3) were obtained from average measured soil water matric head data of the 82.5 cm and 97.5 cm soil depth, with pressure transducer readings corrected for temperature. The soil water dynamics was studied for two irrigation periods. In the first period (8/18–8/25/95), the experimental plot was irrigated three times (8/18, 8/21, and 8/23). The corresponding applied water amounts to the 2.25 x 2.25 m experimental area were 125 L (24.7 mm), 85 L (16.8 mm), and 106 L (20.9 mm). Irrigation scheduling was based on California Irrigation Management Information System (CIMIS) data (Snyder et al., 1987; Goldhamer and Snyder, 1989). Additional information on the applied irrigation scheduling can be found in Koumanov et al. (1997). Neutron probe and tensiometer readings were collected before and after each water application, and daily at about 6:00, 10:00, 14:00, and 18:00 hours. In addition, readings were taken nightly after the 8/18 irrigation – at 22:00 and 2:00 hours. During the second period (9/13–9/29), soil water depletion was monitored during a 16-day period with no irrigation. On 9/13, the micro-sprinkler system was used to wet up the entire 90 cm soil profile above field capacity, after which the irrigation was cut off. Neutron probe and tensiometer readings were taken immediately after water applications at 13:00, 15:00, and 18:00 hours, every four hours daily (at 6:00, 10:00, 14:00, and 18:00 hours) from 9/14 – 9/17, and only one time daily at about 10:00 o’clock from 9/18 – 9/29. In both experimental periods, the water balance was computed between each time interval, from 4 D q (x,y,z,t) D q = - + S(x,y,z,t) [4] D t D z where S(x,y,z,t) denotes the sink term representing root water uptake (cm3 cm-3 h-1) from the corresponding depth interval. Hence, root water uptake can be estimated from the water content measurements only, if drainage between soil layers of thickness ∆z and evaporation are negligible. It will be shown, based on experimental results of this study, that such is the case for the four-hour time intervals between measurements. The soil water content and soil water depletion rates were evaluated for the 0–22.5, 22.5–37.5, 37.5–52.5, 52.5–67.5, 67.5–82.5, and 82.5–97.5 cm soil depth intervals, ∆z ,corresponding to the 15, 30, 45, 60, 75, and 90 cm depth measurements with the neutron probe. For the 25 access tubes and six measurement depths, it took about one hour to complete the 150 readings of the experimental plot. The resulting water content changes for any location and selected time interval were generally less than the standard error (SE) of the neutron probe calibration curves (between 0.01 and 0.02 cm3 cm-3). Therefore, one may question the accuracy of the water depletion estimates. However, we assumed that the SE of the calibration curves were systematic in time, so that water content changes at specific depths were considered insignificant only, if the water content changes between two consecutive measurement times was smaller than the precision of the neutron probe. Each neutron probe measurement consisted of 16 sec counting intervals, resulting in a precision of the neutron probe measurement of about 0.0029 cm3 cm-3, assuming that radioactive decay follows a Poisson distribution. Our measurement precision was about equal to the one provided by the manufacturer (CPN Corporation, 1996/1997). In addition to the measurements of the tree root zone domain, a separate infiltration and drainage experiment was conducted between two tree rows without active roots, nearby the experimental plot. After fully wetting this 1m by 1m area, it was covered and water content measurements in the center of this 1 m2 small plot was monitored from neutron probe measurements, to estimate field capacity (FC) in situ (Cassel and Nielsen, 1986). After drainage stopped, the cover was removed, and evaporation measurements were estimated from water balance calculations for both the top 0 - 22.5 cm layer and the 22.5 - 37.5 cm depth interval. Results and discussion First experimental period Most of the variations in soil water content occurred in the upper soil layer, as measured by the 15-cm depth neutron probe readings. For the remaining larger part of the monitored soil volume, measured soil water content fluctuations were smaller than the precision of the neutron probe. Figure 2a-d presents the patterns of irrigation water distribution (Koumanov et al., 1997), which are different for each irrigation event because of the wind drift and the varying pressure in the system. The average wind speed was 2.0 m s-1 on 8/18, 1.6 m s-1 on 8/21, and 2.7 m s-1 on 8/23, and the application rate measured by a flow meter was 2.86 mm h-1, 3.50 mm h-1, and 3.63 mm h-1, respectively (for more detailed discussion see Koumanov et al., 1997). Fig. 3 shows the spatial patterns of soil water content increase immediately after each of the three water applications (Fig 3a-c) with the corresponding soil water depletion between applications during the first experimental period (Fig. 3d-f). Comparison of Figs. 2 and 3 shows that the nonuniform water application patterns are related to the soil moisture and depletion patterns of Fig. 3. For example, both the maxima of water application rate soil moisture increase and 5 depletion for the 8/18 event are located in the southern part of the experimental plot. It also appears that micro-irrigation scheduling, which was based on drip-irrigation (Snyder et al., 1987; Goldhamer and Snyder, 1989), resulted in under-irrigation and inadequate wetting of the soil profile. Consequently, most of the tree root activity, as indicated by temporal and spatial changes in soil water, was also concentrated in the surface soil layer. Dry soil and small soil water matric potential gradients resulted in very low values of vertical water flux across the lower boundary. Between water applications, we estimated that water flow was generally directed upwards (capillary rise), with maximum flux values occasionally at about 0.0002 cm h-1 (Fig. 4), which were considered insignificant when compared with soil water storage changes across the whole soil profile. The experimental data confirm that most of the soil water dynamics during the first period was occurring in the upper 22.5 cm soil layer only. Comparing Figs. 3a-c with 3d-f, it is clear that the zones of the largest soil water depletion coincide with the areas of maximum water storage increase after irrigation corresponding with high water application rates. We noted that there was no water storage increase measured around the zones of soil water depletion, i.e. lateral water redistribution after water applications was insignificant and could not explain the relatively high values of soil water content uniformity over the experimental plot (Christiansen’s coefficient larger than 0.70) reported by Koumanov et al. (1997). We hypothesize that the spatial distribution of root activity is determined by the water application patterns of the sprinkler. Thus, roots proliferate in regions of the largest water content, thereby also dictating the zones of maximum water depletion by root water uptake. The relative uniform water content distributions in the surface layer before each of the three water applications in Fig. 5-a,b and c confirm our hypothesis. In the presented study, soil water depletion was used as an indicator for root water uptake intensity, although soil evaporation likely contributed to surface soil water depletion as well. We estimated the evaporation E (mm day-1) from bare soil after removing the cover of the 1 m2 experimental plot between the tree rows. We considered this to be a valid way of estimating soil evaporation of the 2.25 m x 2.25 m experimental plot between irrigations, as the surface soil water content was mostly about equal or less than FC, even immediately after irrigation. To obtain the rates of soil water depletion (R) for both soil depth intervals, the evapotranspiration values E (mm day-1) were converted to obtain R in the appropriate dimensional units (cm3 cm-3 day-1). The estimated depletion rates are presented on Fig. 6. Except for the very first measurement immediately after uncovering of the soil surface, all R values were less than 0.0005 cm3 cm-3 h-1. Hence, the change of soil water content in the top layer for the four-hour time intervals between measurements was smaller than 0.002 cm3 cm-3, or less than the precision of the neutron probe. Thus, the influence of evaporation on soil water depletion could be assumed insignificant compared to root water uptake when considering the four- hour time intervals between measurements. The estimated daily evaporation rates agree well with those computed by Vrugt et al. (2001) in their two-dimensional numerical modeling of root water uptake based on data from the same experiment, i.e. under identical soil and meteorological conditions. According to the results obtained, the almond tree appeared capable to redirect its root activity towards the zones of the most favorable water regime. The temporal changes in irrigation water application pattern of the sprinkling system had a beneficial effect on the shallow root system, maintaining an active root system over a large part of the experimental plot. However, the time series of water content measurements did not show soil water depletion by root uptake in zones of high water content at the 60-75 cm depth, likely because of the absence of active tree roots at the larger soil depths. Also, soil surface water storage did not change much at the larger radial distances from the tree (Fig. 6 5). Apparently, the active almond tree roots developed preferentially close to the trunk, provided there is sufficient water to meet plant water requirements. Similar results were obtained for drip irrigated peach trees grown in lysimeters (Koumanov et al., 1998; Stoilov et al., 1999). The rate and the spatial distribution of root water uptake varied significantly, depending on soil water availability, the distance from the tree trunk, and the intensity of meteorological factors during the day. As an example, Fig. 7 presents the spatial variation of surface soil water depletion rate (cm3 cm-3 h-1) on 8/19 and 8/20, two days after the first water application, at 10:00, 14:00, and 18:00 o’clock, respectively. The initial locations of maximum water depletion rates by root water uptake were close to the tree (bottom left corner in Fig. 7-a). Later, as evidenced by the water depletion maps in Figs. 7-b, c, d, e and f, the zones of maximum water uptake moved away from the tree trunk, with roots exploring the wetter soil regions. During the periods between water applications, the pattern of soil water depletion resembled concentric circles with the tree trunk at the center and a radius about equal to the vertical projection of the tree crown. As expected, no soil water depletion or root activity was measured in the experimental plot during night periods. Second experimental period During the second experimental period, the soil water status was monitored after fully wetting the root zone. Figure 8 shows the volumetric soil moisture distribution (cm3 cm-3) throughout the soil profile for all 25 measurement points in Fig. 1, immediately after the irrigation at 13:00 on 09/13, and at 10:00 o'clock on 9/14, 9/15, and 9/16. For the majority of points, except for a few at the periphery of the grid (marked with larger symbols), the soil profile was wetted above FC. Thus, when fully wet, the water content distribution is uniform and there should be no water limitations anywhere in the root domain. However, soil wetting above FC is associated with drainage below the root zone and between layers, which may affect the root water uptake estimates. Figures 8-b,c, and d show that, starting from the day after irrigation, the water content throughout the soil profile gradually decreased approaching FC values, confirming that drainage had taken place with changes in soil moisture distributed fairly equally across the soil profile. Figure 9 shows the time change of drainage rates, D (cm h-1), as computed using Eq. [1] and using the tensiometer readings at the bottom of the rooting zone, and the corresponding calculated soil water depletion rates R (cm3 cm3 h-1) for the top 22.5 and 37.5 cm soil layers, respectively, assuming that the daily decrease in soil water storage by drainage is about equal for all soil layers. Drainage rates were mostly smaller than 0.003 cm/h or 0.7 mm/d. We note that the inferred depletion rates are much smaller than presented in Fig. 8 from the soil moisture measurements. Thus, we can assume that the low drainage rates did not affect the root water uptake patterns that developed mostly in the upper 20—25 cm soil layer. Likely, the insufficient soil wetting below the 25 cm soil depth during the irrigation season has resulted in root suberization and reduced development of active roots. The preferential root development in the upper soil layers under microsprinkling has been reported before (Dasberg et al., 1985; Hamer, 1987; Kjelgren et al., 1985; Meyer and Peck, 1985; Roth and Gardner, 1985). As demonstrated in Fig. 10-a and b, root water uptake developed initially close to the tree trunk, shifting to wetter parts of the root zone as water becomes depleted there, Figures 10-c,d,e and f. Also Green et al. (1997) and Andreu et al. (1997) have reported on temporal changes of maximum root water uptake patterns, as determined by variations in water availability. 7 Throughout the first few days of the second experimental period, water uptake rate was mostly controlled by meteorological conditions such as temperature, air humidity, and wind; daily maximum values occurred between 10:00 and 18:00 o’clock and negligible uptake was observed during the night. In the lateral direction, the pattern of soil water content decrease radially outwards from the tree trunk, with the radial distance controlled by the vertical projection of the tree crown, as demonstrated in Fig. 5-d for water content distribution on 9/21. In the vertical direction, water depletion started at the soil surface and progressed downwards with time. However, daily maximum root uptakes decreased with time as caused by the reduced water content in the upper soil layers with the active roots, and the absence of roots in the lower soil layers despite that water content was relatively high there. As expected, the maximum values of total soil water depletion were found at the soil surface. We expect that part of the total soil water depletion was caused by soil evaporation, yet because of the sandy soil material evaporation plays a minor role as the soil surface becomes dry. Evaporation rates were not measured, though there is no evidence that it would affect the spatial root water uptake patterns. The lesser role of evaporation was also confirmed by the low rates of soil water depletion in surface soil zones with low root activity. Root water uptake rate depends on root density; however, fluxes into the roots are directly controlled by the gradient in water potential between the roots and the surrounding soil. In the range of soil water potential where the root resistance is the limiting factor controlling water uptake, maximum water uptake will occur in the zones of maximum soil water potential or water content (Veihmeyer and Hendrickson, 1938; Hough et al., 1965). Following this same reasoning, it is expected that the tree root water potential increases with increasing distance of the root from the tree trunk, thereby causing a radial spatial root water uptake distribution. Root water potential increases also in depth because of both the distance and the gravity. Hence, we postulate that in a uniformly-wetted soil, root extraction will be higher near the trunk and close to soil surface, because of the larger water potential gradient there. In summary then, the spatial distribution of root water uptake S (x, y, z) may be expressed as a function of root density (d), soil water potential, (h ) and root water potential (h ), i.e. S (x, y, z) = f (d, h , h ), where h = f (l, γ), with l s r s r r denoting the distance from the tree trunk and γ denoting the depth. Although the experimental results are solely for a single almond tree, it is expected that the established pattern of root water uptake are likely for most other micro-sprinkler irrigated fruit tree species. Conclusions Regions of high soil water depletion by root water uptake coincided with areas of high local rates of water application as caused by the nonuniform water application by the microsprinkler system. Despite that water application was nonuniform, soil moisture uniformity prior to irrigation was large, and was caused by differential root water uptake in the surface soil. Throughout the experiments, the roots of the almond tree were capable to redirect their areas of maximum root activity towards the zones of the most favorable water regime, thereby resulting in fairly uniform water content distributions. Typically, zones of maximum root water uptake developed from the tree trunk towards the outer regions of the root zone, shifting to wetter parts of the root zone domain. Consequently, soil water depletion patterns formed a radial pattern around the tree trunk. We did not measure much root activity below the top 30 cm, likely because of limited soil 8 wetting below this depth. As a result, there was probably little root development below the soil surface layer and soil water depletion was small even if the total 90-cm soil profile was wetted. Thus, in summary, factors controlling root water uptake in irrigated tree crops are (1) spatial distribution of active roots, (2) root zone water content distribution, and (3) distance from the tree trunk. Obtained results show that water use efficiency could be increased if irrigation water was applied in a circular pattern with decreasing water application with increasing distance of the tree trunk. Such a surface water application pattern can be achieved by a pair of sector-operating microsprinklers (e.g. 210° each), located at both sides of the tree trunk. Acknowledgements This research was supported in part by Grant IS-2131-92RC from the U.S.-Israel Binational Agricultural Research and Development fund (BARD). We also thank the International Atomic Energy Agency (IAEA) for the funding for Dr. Koumanov's fellowship in the Department of Land, Air and Water Resources at the University of California, Davis for the June 1995 – March 1996 period. 9 References Andreu L, Hopmans JW, Schwankl LJ (1997) Spatial and temporal distribution of soil water balance for a drip-irrigated almond tree. Agricultural Water Management 35: 123-146. Behboudian MH, Dixon J, Pothamshetty K (1998) Plant and fruit responses of lysimeter-grown “Breeburn” apple to deficit irrigation. Journal of Horticultural Science & Biotechnology 73(6): 781-785. Bielorai H (1985) Moisture, salinity and root distribution of drip irrigated grapefruit. Proc. 3rd Int. Drip/Trickle Irrigation Congr., Fresno, Vol 2: 562-567. Bucks DA, Davis S (1986) Introduction. In: Nakayama FS, Bucks DA (eds) Trickle Irrigation for Crop Production. Elsevier Science Publishers B.V., pp 1-26. Cassel DK, Nielsen DR (1986) Field capacity and available water capacity. In: Klute A (ed.) Methods of Soil Analysis, Part 1 Physical and Mineralogical Methods. SSSA Book Series: 5, pp 901-924. Chalmers DJ Mitchell PD, Jerie PH (1984) The physiology of peach and pear trees using reduced irrigation. Acta Hort. 146: 143-149. Clausnitzer V, Hopmans JW (1994) Simultaneous modeling of transient three- dimensional root growth and soil water flow. Plant and Soil 164: 299-314. Clothier BE (1989) Research imperatives for irrigation science. Journa of Irrigation and Drainage Engineering 115(3): 421-448. Coelho FE, Or D (1996) A parametric model for two-dimensional water uptake intensity by corn roots under drip irrigation. Soil. Sci. Soc Am. J. 60: 1039-1049. CPN Corporation (1986/1987) 503 HYDROPROBE for Irrigation Water Management. Form No 503186BR, 4 pp. Cross JV, Dickler E (eds) (1994) Guidelines for Integrated Production of Pome Fruits in Europe, Technical Guideline III, 2nd Edition. IOBC wprs Bulletin 17(9), 8 pp. Dasberg S, Embleton TW, Bielorai H, Erner Y, Stolzy LH (1985) Irrigation and fertilizer management of citrus for optimum reduction and minimum pollution. Final Report, Institute of Soils and Water, ARO, Bet Dagan, Israel. Goldhamer DA, Snyder RL (1989) Irrigation Scheduling. A guide for efficient on- farm water management. University of California, DANR Publication 21454, 67 pp. Goldhamer DA (1996) Regulated deficit irrigation of fruit and nut trees. 7 th International Conference on Water and Irrigation, Tel Aviv, Israel, pp.152-167. Green SR, Clothier BE, McLeod DJ (1997) The response of sap flow in apple roots to localized irrigation. Agricultural Water Management 33(1): 63-78. Hamer, PJC (1987) Irrigation of fruit: 3. Irrigation methods. Irrigation News 11: 23-28. Harradine FF (1948) Soils of Colusa county. Report. University of California Internal Report. University of California, Davis, CA 95616. Kjelgren R, Goldhamer DA, Uriu K, Weinbaum SA (1985) Almond tree response to variable nitrogen fertilization rates through low volume sprinklers. Proc. 3rd Int. Drip/Trickle Irrigation Congr., Fresno, Vol 2. Koumanov K, Dochev D, Stoilov G (1998) Investigations on fertigation of peach on three soil types - patterns of soil wetting. Bulg. J. Agric. Sci. 4: 745-753. Koumanov KS, Hopmans JW, Schwankl LW, Andreu L, Tuly A (1997) Application efficiency of micro-sprinkler irrigation of almond trees. Agricultural water Management 34: 247-263. Kramer PJ, Boyer JS (1995) Water Relations of Plants and Soils. Academic Press, San Diego, 495 pp. 10
Description: